September 20, 2021

How can you ensure batch-to-batch consistency when roasting coffee?


Consistency in coffee roasting is something that’s often discussed when things go wrong, rather than right. 

For customers buying roasted coffee, consistency is an expectation, not a benefit. People expect their roasters to reliably reproduce the same flavour profiles again and again. 

But how do you guarantee consistency? And which variables should you keep an eye on? Read on to answer these questions and learn more. 

You might also like our article on automation in coffee roasting.

Coffee beans exit a coffee roaster that roasts for batch-to-batch consistency.

How do we define consistency in coffee roasting?

The science of coffee roasting comes down to slight tweaks to heat and airflow during the roast. Increasing or reducing these factors at different points allows roasters to manipulate the innate flavour locked within green coffee. 

But once a roast profile has been developed, there is another challenge: repeating it. 

This is where the question of consistency comes into play. We can consider roasters to be consistent when each bag of coffee tastes the same as the last. There should be no fluctuation from batch to batch, and every customer who buys the same coffee should broadly experience similar flavours (albeit with some variation for palate).

This is important, as consistently delivering a high-quality product is the foundation of a specialty coffee roaster’s brand. It’s also one of the most important steps to maintaining a loyal customer base.

A roaster observes the roast curve.

Measuring consistency: Data & colour

Morten Münchow is a coffee roaster, trainer, and researcher at Coffee Mind in Denmark. He explains that there are several variables that can affect consistency from batch to batch. Conversely, there are a number of different ways to measure consistency using data from your roaster. 

Morten tells me about a study that Coffee Mind conducted in partnership with Stronghold, a roaster manufacturer based in South Korea. 

The study, he says, pinpointed several different data points that roasters can use to measure consistency from roast to roast. They are:

  • Exact time and temperature of first crack
  • Temperature of the room
  • End temperature time of the bean
  • Weight loss during the roast
  • Agtron or ColorTrack measurements

The last of these, Agtron or ColorTrack measurements, are popular among roasters for measuring “colour”. Agtron or ColorTrack equipment shines a light on a sample of coffee beans, before providing a reading based on the colour (opacity) of the sample used. 

The higher the number, the lighter the roast. An Agtron “score” of around 70, for instance, indicates what would typically be considered a light roast. Medium roasts are around 55 to 60, while dark roasts start at 35 and can drop as low as 25.

In the study, Morten says Coffee Mind and Stronghold roasted several batches using a Stronghold S7X. The aim, he says, was to achieve similar readings for each batch while using the same time and heat parameters. 

The difference in consistency was measured using the average difference in Agtron number from batch to batch.

“After the first roast of the day, we were able to reproduce the same roast profile with an average deviation of 0.7 Agtron,” Morten explains.

Coffee beans exit a coffee roaster.

Which variables can affect roast consistency?

Unfortunately, unless your roastery is maintained with scientific precision, environmental variables and human error will naturally affect roast consistency. 

Furthermore, while highly-skilled roasters are able to monitor and adjust a roast profile if they see an issue, responding to these variables takes years of experience and understanding.

Morten tells me that the biggest variable that affects consistency in batch roasting is heat retention.

“Any retained heat in the material of the roaster can make the beans behave differently,” he says. “This can be the case even if your input parameters are the same (heat, rotation speed, airflow, and so on).”

Consequently, this means that any potential heat transfer into and out of your roaster can affect consistency – even the ambient temperature of your roastery.

For instance, in the summer months, a roaster may need less time to warm up than it would in the winter. Consequently, on colder days, the roaster can take more time to preheat to the right temperature. For inexperienced roasters, a lack of knowledge about how ambient temperature affects the roast can mean inconsistency from the first batch of the day to the last. 

“Almost any roasting professional knows that the first roast of the day is an ‘off roast’,” Morten says. “This is an issue for almost all roaster manufacturers. It often requires the roastmaster to adjust the profile in order to hit the target colour and Agtron reading. 

“The underlying reason for this is usually that the roaster is not sufficiently heated by its pre-heating sequence alone.”

Morten also notes that batch size affects consistency. “I don’t even consider it a potential inconsistency,” he says. “It’s such a direct and predictable source of inconsistency, that I think you just have to dial in a new profile for each batch size.”

However, if you are changing batch size and want to measure consistency, Morten recommends using infrared (IR) temperature probing.

“On the Stronghold S7X and other roasters using infrared temperature probes, the results on measuring different batch sizes have actually been really good,” he says. “These actually measure the surface of the coffee without being affected by the surroundings – which bean temperature probes can be.

“[If you are going to change batch size], the infrared probe is a good technological way to manage consistency,” he explains. “However, this is only as long as you’re working with the same coffee on the same roaster.”

A countertop coffee roaster alongside a coffee grinder.

How can we minimise inconsistency?

The first step is simply to develop and refine roast profiles, and follow these to the letter for every batch. However, as mentioned previously, human error can be a factor, especially if you have larger teams working on one roaster.

Morten recommends setting out a rigorous routine and trying to isolate yourself as best as possible.

“Plan the day before you start roasting, and make sure nobody disturbs you,” Morten says. “You need to make sure you’re doing nothing but roasting for that planned time. No phone, no emails, no social media!”

The other solution he notes is the use of technology. While an experienced roastmaster’s sight and smell alone can improve consistency, modern roasters are naturally more precise and can pull on a range of data points.

Morten says the study Coffee Mind conducted with Stronghold found that the Stronghold S7X’s auto-replication mode can replicate the precision of even the most trained roasters.

“The Stronghold S7X roaster has proven to be capable of a precise reproduction of coffee roast profiles,” he says. “[An average deviation of 0.7] means the S7X’s auto-replication mode is as precise as highly trained individuals on manual roasters.

“It’s difficult to imagine what you could even do wrong with the S7X other than adding the wrong amount of beans to the hopper,” he says. “It also auto-dumps the coffee at the end of the roast, making the last and most crucial decision – when to finish the roast – more consistent.”

This new, technological approach to roasting may be difficult to adopt for those who have been trained to roast the “traditional” way. However, being able to roast without a need for constant supervision means roasters can spend more time on other areas of the business.

A woman operates a coffee roaster.

How else can roasters make their batches more consistent?

Other than developing a solid profile and replicating it to the best of your ability, Morten says that using and analysing data is a great way to be more consistent. 

Time, end temperature, and weight loss are all good places to start. However, Morten notes that recording different scenarios across each roaster you use can help you analyse any gaps in consistency. 

“Test consistency with your roaster and explore the different scenarios that could lead to inconsistency,” he recommends. “The key parameter for consistency is flavour, which is 80% driven by variation in colour and 20% driven by variation in timing.”

He says the study conducted by Coffee Mind and Stronghold found that the specific protocol followed for each batch made the largest difference to consistency.

“It doesn’t happen enough, but roaster manufacturers’ user manuals should really specify how the operator should handle the roaster to achieve peak consistency,” he adds.

Some of the measures Morten recommends include:

  • Waiting the same amount of time between each batch (to ensure uniformity in heat transfer)
  • Extended pre-heating where necessary
  • Rapid pre-heating after finishing a batch
  • Roasting the first batch manually to minimise the impact of starting with a cold roaster
A professional coffee roaster observes the roast curve to ensure they are roasting for consistency.

It’s clear that consistency is a key parameter for roasters to consider when working with larger volumes of coffee. After dialling in that perfect roast profile, it’s of the utmost importance that you replicate it perfectly to make sure customers keep coming back.

Understanding how heat transfers into and out of your roaster is a great place to start, as is drawing on all the data points you have available. However, going forward, it seems like there will be no substitute for the precision that a modern roaster can offer – no matter how experienced a roastmaster is, human error will always be a factor. 

Enjoyed this? Then try our article on choosing a roaster for your coffee shop.

Photo credits: Café Muda, Monogram Coffee

Perfect Daily Grind

Please note: Stronghold is a sponsor of Perfect Daily Grind.

Want to read more articles like this? Sign up for our newsletter!